You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

403 lines
13 KiB

/*
* NuppelVideo demuxer.
* Copyright (c) 2006 Reimar Doeffinger
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "libavutil/channel_layout.h"
#include "libavutil/imgutils.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/intfloat.h"
#include "avformat.h"
#include "internal.h"
#include "riff.h"
static const AVCodecTag nuv_audio_tags[] = {
{ AV_CODEC_ID_PCM_S16LE, MKTAG('R', 'A', 'W', 'A') },
{ AV_CODEC_ID_MP3, MKTAG('L', 'A', 'M', 'E') },
{ AV_CODEC_ID_NONE, 0 },
};
typedef struct NUVContext {
int v_id;
int a_id;
int rtjpg_video;
} NUVContext;
typedef enum {
NUV_VIDEO = 'V',
NUV_EXTRADATA = 'D',
NUV_AUDIO = 'A',
NUV_SEEKP = 'R',
NUV_MYTHEXT = 'X'
} nuv_frametype;
static int nuv_probe(const AVProbeData *p)
{
if (!memcmp(p->buf, "NuppelVideo", 12))
return AVPROBE_SCORE_MAX;
if (!memcmp(p->buf, "MythTVVideo", 12))
return AVPROBE_SCORE_MAX;
return 0;
}
/// little macro to sanitize packet size
#define PKTSIZE(s) (s & 0xffffff)
/**
* @brief read until we found all data needed for decoding
* @param vst video stream of which to change parameters
* @param ast video stream of which to change parameters
* @param myth set if this is a MythTVVideo format file
* @return 0 or AVERROR code
*/
static int get_codec_data(AVFormatContext *s, AVIOContext *pb, AVStream *vst,
AVStream *ast, int myth)
{
nuv_frametype frametype;
if (!vst && !myth)
return 1; // no codec data needed
while (!avio_feof(pb)) {
int size, subtype, ret;
frametype = avio_r8(pb);
switch (frametype) {
case NUV_EXTRADATA:
subtype = avio_r8(pb);
avio_skip(pb, 6);
size = PKTSIZE(avio_rl32(pb));
if (vst && subtype == 'R') {
if ((ret = ff_get_extradata(NULL, vst->codecpar, pb, size)) < 0)
return ret;
size = 0;
if (!myth)
return 0;
}
break;
case NUV_MYTHEXT:
avio_skip(pb, 7);
size = PKTSIZE(avio_rl32(pb));
if (size != 128 * 4)
break;
avio_rl32(pb); // version
if (vst) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
vst->codecpar->codec_tag = avio_rl32(pb);
vst->codecpar->codec_id =
ff_codec_get_id(ff_codec_bmp_tags, vst->codecpar->codec_tag);
if (vst->codecpar->codec_tag == MKTAG('R', 'J', 'P', 'G'))
vst->codecpar->codec_id = AV_CODEC_ID_NUV;
} else
avio_skip(pb, 4);
if (ast) {
int id;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
ast->codecpar->codec_tag = avio_rl32(pb);
ast->codecpar->sample_rate = avio_rl32(pb);
if (ast->codecpar->sample_rate <= 0) {
av_log(s, AV_LOG_ERROR, "Invalid sample rate %d\n", ast->codecpar->sample_rate);
return AVERROR_INVALIDDATA;
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
ast->codecpar->bits_per_coded_sample = avio_rl32(pb);
ast->codecpar->channels = avio_rl32(pb);
ast->codecpar->channel_layout = 0;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
id = ff_wav_codec_get_id(ast->codecpar->codec_tag,
ast->codecpar->bits_per_coded_sample);
if (id == AV_CODEC_ID_NONE) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
id = ff_codec_get_id(nuv_audio_tags, ast->codecpar->codec_tag);
if (id == AV_CODEC_ID_PCM_S16LE)
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
id = ff_get_pcm_codec_id(ast->codecpar->bits_per_coded_sample,
0, 0, ~1);
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
ast->codecpar->codec_id = id;
ast->need_parsing = AVSTREAM_PARSE_FULL;
} else
avio_skip(pb, 4 * 4);
size -= 6 * 4;
avio_skip(pb, size);
return 0;
case NUV_SEEKP:
size = 11;
break;
default:
avio_skip(pb, 7);
size = PKTSIZE(avio_rl32(pb));
break;
}
avio_skip(pb, size);
}
return 0;
}
static int nuv_header(AVFormatContext *s)
{
NUVContext *ctx = s->priv_data;
AVIOContext *pb = s->pb;
char id_string[12];
double aspect, fps;
int is_mythtv, width, height, v_packs, a_packs, ret;
AVStream *vst = NULL, *ast = NULL;
avio_read(pb, id_string, 12);
is_mythtv = !memcmp(id_string, "MythTVVideo", 12);
avio_skip(pb, 5); // version string
avio_skip(pb, 3); // padding
width = avio_rl32(pb);
height = avio_rl32(pb);
avio_rl32(pb); // unused, "desiredwidth"
avio_rl32(pb); // unused, "desiredheight"
avio_r8(pb); // 'P' == progressive, 'I' == interlaced
avio_skip(pb, 3); // padding
aspect = av_int2double(avio_rl64(pb));
if (aspect > 0.9999 && aspect < 1.0001)
aspect = 4.0 / 3.0;
fps = av_int2double(avio_rl64(pb));
if (fps < 0.0f) {
if (s->error_recognition & AV_EF_EXPLODE) {
av_log(s, AV_LOG_ERROR, "Invalid frame rate %f\n", fps);
return AVERROR_INVALIDDATA;
} else {
av_log(s, AV_LOG_WARNING, "Invalid frame rate %f, setting to 0.\n", fps);
fps = 0.0f;
}
}
// number of packets per stream type, -1 means unknown, e.g. streaming
v_packs = avio_rl32(pb);
a_packs = avio_rl32(pb);
avio_rl32(pb); // text
avio_rl32(pb); // keyframe distance (?)
if (v_packs) {
vst = avformat_new_stream(s, NULL);
if (!vst)
return AVERROR(ENOMEM);
ctx->v_id = vst->index;
ret = av_image_check_size(width, height, 0, s);
if (ret < 0)
return ret;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
vst->codecpar->codec_type = AVMEDIA_TYPE_VIDEO;
vst->codecpar->codec_id = AV_CODEC_ID_NUV;
vst->codecpar->width = width;
vst->codecpar->height = height;
vst->codecpar->bits_per_coded_sample = 10;
vst->sample_aspect_ratio = av_d2q(aspect * height / width,
10000);
#if FF_API_R_FRAME_RATE
vst->r_frame_rate =
#endif
vst->avg_frame_rate = av_d2q(fps, 60000);
avpriv_set_pts_info(vst, 32, 1, 1000);
} else
ctx->v_id = -1;
if (a_packs) {
ast = avformat_new_stream(s, NULL);
if (!ast)
return AVERROR(ENOMEM);
ctx->a_id = ast->index;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
ast->codecpar->codec_type = AVMEDIA_TYPE_AUDIO;
ast->codecpar->codec_id = AV_CODEC_ID_PCM_S16LE;
ast->codecpar->channels = 2;
ast->codecpar->channel_layout = AV_CH_LAYOUT_STEREO;
ast->codecpar->sample_rate = 44100;
ast->codecpar->bit_rate = 2 * 2 * 44100 * 8;
ast->codecpar->block_align = 2 * 2;
ast->codecpar->bits_per_coded_sample = 16;
avpriv_set_pts_info(ast, 32, 1, 1000);
} else
ctx->a_id = -1;
if ((ret = get_codec_data(s, pb, vst, ast, is_mythtv)) < 0)
return ret;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
ctx->rtjpg_video = vst && vst->codecpar->codec_id == AV_CODEC_ID_NUV;
return 0;
}
#define HDRSIZE 12
static int nuv_packet(AVFormatContext *s, AVPacket *pkt)
{
NUVContext *ctx = s->priv_data;
AVIOContext *pb = s->pb;
uint8_t hdr[HDRSIZE];
nuv_frametype frametype;
int ret, size;
while (!avio_feof(pb)) {
int copyhdrsize = ctx->rtjpg_video ? HDRSIZE : 0;
uint64_t pos = avio_tell(pb);
ret = avio_read(pb, hdr, HDRSIZE);
if (ret < HDRSIZE)
return ret < 0 ? ret : AVERROR(EIO);
frametype = hdr[0];
size = PKTSIZE(AV_RL32(&hdr[8]));
switch (frametype) {
case NUV_EXTRADATA:
if (!ctx->rtjpg_video) {
avio_skip(pb, size);
break;
}
case NUV_VIDEO:
if (ctx->v_id < 0) {
av_log(s, AV_LOG_ERROR, "Video packet in file without video stream!\n");
avio_skip(pb, size);
break;
}
ret = av_new_packet(pkt, copyhdrsize + size);
if (ret < 0)
return ret;
pkt->pos = pos;
pkt->flags |= hdr[2] == 0 ? AV_PKT_FLAG_KEY : 0;
pkt->pts = AV_RL32(&hdr[4]);
pkt->stream_index = ctx->v_id;
memcpy(pkt->data, hdr, copyhdrsize);
ret = avio_read(pb, pkt->data + copyhdrsize, size);
if (ret < 0) {
return ret;
}
if (ret < size)
av_shrink_packet(pkt, copyhdrsize + ret);
return 0;
case NUV_AUDIO:
if (ctx->a_id < 0) {
av_log(s, AV_LOG_ERROR, "Audio packet in file without audio stream!\n");
avio_skip(pb, size);
break;
}
ret = av_get_packet(pb, pkt, size);
pkt->flags |= AV_PKT_FLAG_KEY;
pkt->pos = pos;
pkt->pts = AV_RL32(&hdr[4]);
pkt->stream_index = ctx->a_id;
if (ret < 0)
return ret;
return 0;
case NUV_SEEKP:
// contains no data, size value is invalid
break;
default:
avio_skip(pb, size);
break;
}
}
return AVERROR(EIO);
}
/**
* \brief looks for the string RTjjjjjjjjjj in the stream too resync reading
* \return 1 if the syncword is found 0 otherwise.
*/
static int nuv_resync(AVFormatContext *s, int64_t pos_limit) {
AVIOContext *pb = s->pb;
uint32_t tag = 0;
while(!avio_feof(pb) && avio_tell(pb) < pos_limit) {
tag = (tag << 8) | avio_r8(pb);
if (tag == MKBETAG('R','T','j','j') &&
(tag = avio_rb32(pb)) == MKBETAG('j','j','j','j') &&
(tag = avio_rb32(pb)) == MKBETAG('j','j','j','j'))
return 1;
}
return 0;
}
/**
* \brief attempts to read a timestamp from stream at the given stream position
* \return timestamp if successful and AV_NOPTS_VALUE if failure
*/
static int64_t nuv_read_dts(AVFormatContext *s, int stream_index,
int64_t *ppos, int64_t pos_limit)
{
NUVContext *ctx = s->priv_data;
AVIOContext *pb = s->pb;
uint8_t hdr[HDRSIZE];
nuv_frametype frametype;
int size, key, idx;
int64_t pos, dts;
if (avio_seek(pb, *ppos, SEEK_SET) < 0)
return AV_NOPTS_VALUE;
if (!nuv_resync(s, pos_limit))
return AV_NOPTS_VALUE;
while (!avio_feof(pb) && avio_tell(pb) < pos_limit) {
if (avio_read(pb, hdr, HDRSIZE) < HDRSIZE)
return AV_NOPTS_VALUE;
frametype = hdr[0];
size = PKTSIZE(AV_RL32(&hdr[8]));
switch (frametype) {
case NUV_SEEKP:
break;
case NUV_AUDIO:
case NUV_VIDEO:
if (frametype == NUV_VIDEO) {
idx = ctx->v_id;
key = hdr[2] == 0;
} else {
idx = ctx->a_id;
key = 1;
}
if (stream_index == idx) {
pos = avio_tell(s->pb) - HDRSIZE;
dts = AV_RL32(&hdr[4]);
// TODO - add general support in av_gen_search, so it adds positions after reading timestamps
av_add_index_entry(s->streams[stream_index], pos, dts, size + HDRSIZE, 0,
key ? AVINDEX_KEYFRAME : 0);
*ppos = pos;
return dts;
}
default:
avio_skip(pb, size);
break;
}
}
return AV_NOPTS_VALUE;
}
AVInputFormat ff_nuv_demuxer = {
.name = "nuv",
.long_name = NULL_IF_CONFIG_SMALL("NuppelVideo"),
.priv_data_size = sizeof(NUVContext),
.read_probe = nuv_probe,
.read_header = nuv_header,
.read_packet = nuv_packet,
.read_timestamp = nuv_read_dts,
.flags = AVFMT_GENERIC_INDEX,
};